SwiftKey is one of my favourite keyboards and I've been a long time supporter of the business. Their keyboard algorithms greatly increased my typing speed on a smartphone and I'm sure have inspired other keyboard manufacturers, and Google themselves, to work on sharpening up their own keyboard designs. Today, however, SwiftKey have announced that they have been working with both Intel and Professor Stephen Hawking to design and build a custom keyboard to make life easier for the well known scientist. SwiftKey have integrated their technology into Professor Hawking's communication system and it has approximately doubled the speed system. Professor Hawking makes typing mistakes very differently to most smartphone users; he is an accurate and precise speller, whereas most people (myself included) have fast, wildly inaccurate and yet still-correctly guessed typing styles! SwiftKey's technology is making it much easier for Professor Hawking to continue writing long-hand lectures and of course, books.
The core of improving the language system is that SwiftKey better understand how the Professor uses language. Stephen's communication system is based around a small sensor activated by a muscle in his cheek, which he uses to type characters and numbers onto a keyboard. SwiftKey has built their prediction engine into the software, such that it can now accurately predict whole words (not just characters). Of course, as anybody who has used SwiftKey before will know, the more you use it, the more accurate and faster it becomes. SwiftKey also learns by studying the language we use based on our text messages, Gmail and social networking. In the case of Professor Stephen Hawking, the SwiftKey built a language model based on Professor Hawking's extensive works, both published and unpublished. Essentially, the system was pre-configured to suit the Professor. The system continued to refine its understanding of how the Professor communicates; the team was aiming to produce a system that enables him to write and speak for longer periods of time.
The SwiftKey project leader, Joe Osborne, said: "It has been a wonderful experience to work so closely with Professor Hawking and his team on this project, integrating our technology into his communications system. SwiftKey's predictive technology learns from each individual user, gaining insights from their personal language style". SwiftKey for Android has recently been updated with performance improvements after the interface was revised with Material Design influences. If you've not tried the app, I would recommend going ahead and giving it a try. You might come away mightily impressed!